278 University of Kansas Geological Survey. 
would be left behind farther down stream where surface action 
began. 
A third factor in the production of sand dunes is the action of 
wind. The same winds that separated the silt and clay particles 
from the sand and blew them to great distances producing by their 
lodgment the plains marl likewise moved the sand grains from 
place to place and caused them to accumulate in certain areas 
which we now call sand hills, or sand dunes. The separating process 
of the wind is similar in most respects to that of water, so that here 
the sand left behind is a residual product from which the finer 
particles of clay and silt have been removed. The wind is con- 
stantly working over the loose surface materials, separating them 
into classes dependent upon their facility for being moved, and is 
producing accumulations of one kind at one place by making ad- 
ditions, and of another kind elsewhere by having matter taken 
away from it. 
It is believed that all or nearly all the sand hills of the entire 
state have been produced by one or another of these agents or by a 
combination of them. It is not infrequent to find gravel near the top 
of the sand hills on the south side of the Arkansas river, gravels 
which are too large to have been blown out of the river and into 
their present lodgment places. The idea! that the sand hills on the 
south side of the Arkansas are due to northerly winds blowing sand 
from the river, as the same is now being carried downward from year 
to year, is hardly compatible with the presence of the larger gravel 
in the sandhills as just mentioned. Neither is it capable of ex- 
plaining the occurrence of so many sandy areas in other parts of 
the state. If a north wind is more prevalent in the vicinity of the 
Arkansas river, so as to blow the sand from the river southward 
rather than northward, it may well be asked why we find so much 
sand north of the Cimarron river in the vicinity of Englewood and 
elsewhere, as has already been stated. The two localities are so 
close together that of course when a strong north wind prevailed 
at one place a correspondingly strong one would prevail at the 
other. It is by no means contended that the wind has not blown 
1G. K. Gilbert, Underground Water of the Arkansas Valley in Hastern Colorado, 
Dp. cues Extracted from the Seventeenth Annual Report of the Director U. S. Geologi- 
cal Survey. 
